I respect Stanton Peele, if for no other reason than simply because he is well informed and doesn't mind telling us all about the way he sees things. However, even the mighty sometime misstep, and this seems to be one of those cases.
In his recent post about why the disease concept of alcoholism, or addiction, is bull$&%# (his words) Dr. Peele decides to quote a piece from the NIAAA's website that states that approximately 75% of people who've met the definition of alcohol dependence in their lifetime quit by themselves without any outside intervention. That's great, but what he forgot to also quote is another passage that states that while "70 percent of [alcoholics] have a single episode of less than 4 years, the remainder experience an average of five episodes. Thus, it appears that there are two forms of alcohol dependence: time-limited, and recurrent or chronic."
Maybe Stanton missed this sentence since it was a few lines above the one he was focusing on, but what it's telling us is that the vast majority of people who meet alcohol dependence criteria do so for a very limited amount of time (seemingly their 4 years of college) while another 30% or so (or 25% according to the line Dr. Peele decided to use) have the chronic-relapsing version of alcoholism we've all come to know. So yes, most people quit without help, but the rest have a hell of a time quitting and most of them need help and even then don't necessarily respond to addiction treatment. I don't know that this is very different from the percentage of people that eat too much and gain weight—some stop and return to a normal BMI, the rest become obese. The same story holds for the prediabetics who never quite cross that line but once they do, will need insulin and a strictly managed diet. In both cases I don't think we need to discount the latter because the former exists.